From antoine at devloop.org.uk Wed Apr 1 16:38:18 2020 From: antoine at devloop.org.uk (Antoine Martin) Date: Wed, 1 Apr 2020 22:38:18 +0700 Subject: [winswitch] [ANNOUNCE] Xpra 3.0.8: nothing major In-Reply-To: References: Message-ID: <59185b49-771a-93c8-e9f9-c0130e305d26@devloop.org.uk> Hi, I managed to mess up the very last patch to go in this 3.0.8 release, which caused a regression for those connecting to 3.0.8 servers via ssh (system ssh, xpra's builtin ssh server was not affected). A patch has been applied and I have re-issued the builds for all the affected platforms. (that's all of them except MS Windows) Sorry about the hassle. Cheers, Antoine On 31/03/2020 18:40, Antoine Martin wrote: > Hi, > > There's a fair number of bug fixes in this release, but none of them are > critical. > The clipboard and keyboard (AltGr) handling are the most important ones > in there, the rest is all fairly minor stuff. (platforms, warnings, etc) > > Release notes: > * handling of dpi command line switch (correctly this time?) > * bug report window not getting focus on MacOS > * spurious ssh key warnings with newer versions of paramiko > * AltGr mode with non-X11 clients, layout-group changes > * rare unexpected client exit on MS Windows > * MS Windows clipboard: > update failures > convert CRLF line endings > * Clipboard: > fix selection not shown as active in menus (MS Windows and MacOS) > fix spurious warnings when sharing a session > fix clipboard reset with python2 builds > selection translation for outbound data > support client applications that don't use TARGETS (ie: Motif) > reject invalid targets > * 'xpra upgrade' wrongly updating non-xpra displays > * logging error in client geometry debugging output > * pulseaudio start command with Ubuntu Xenial > * spurious de-iconifications > * handling of server control commands with python3 clients > * UDP backport bug sending control packets > * vfb getting killed on upgrade failure > * proxy server cleanup: force forwarders to terminate > * session info errors during client exit > * printer cleanup errors with invalid UTF8 printer names > * transient-for popup window workaround > * missing libyuv csc module on MacOS > * unicode errors saving xpra runner shell script with python3 > * try to prevent conflicts with Fedora's packages > * better detection of Wayland environments > * use python3 (if installed) by default on Ubuntu Xenial > * don't use Xdummy on arm, too slow > * don't show 'Download' button that we can't honour > * show all pressed keys according to X11 server in 'xpra info' > * try harder not to use video for tiny areas > > Cheers, > Antoine > From crl.langlois at gmail.com Wed Apr 1 16:39:29 2020 From: crl.langlois at gmail.com (carl langlois) Date: Wed, 1 Apr 2020 11:39:29 -0400 Subject: [winswitch] Keypads number output wring value. Message-ID: Hi, I am having issue with the keypad on my keyboard. When pressing number the character display is not the number . i.e when i press 5 i got a E in the terminal. I have enable the debug keyboard logging. When i press 5 on the keypad the verbose output this. On the client side : 2020-04-01 11:33:58,596 parse_key_event(, True)= 2020-04-01 11:33:58,639 handle_key_action(ClientWindow(16), ) wid=16 2020-04-01 11:33:58,670 key_handled_as_shortcut: shortcut(KP_5)=None 2020-04-01 11:33:58,672 send_delayed_key() delayed_event=None 2020-04-01 11:33:58,674 send_key_action(16, ) 2020-04-01 11:33:58,706 mask_to_names() GetKeyState(VK_NUMLOCK)=1, names=['mod2'] 2020-04-01 11:33:58,708 parse_key_event(, False)= 2020-04-01 11:33:58,735 handle_key_action(ClientWindow(16), ) wid=16 2020-04-01 11:33:58,739 key_handled_as_shortcut: shortcut(KP_5)=None 2020-04-01 11:33:58,741 send_delayed_key() delayed_event=None 2020-04-01 11:33:58,774 send_key_action(16, ) 2020-04-01 11:33:59,201 mask_to_names() GetKeyState(VK_NUMLOCK)=1, names=['mod2'] 2020-04-01 11:33:59,208 mask_to_names() GetKeyState(VK_NUMLOCK)=1, names=['mod2'] 2020-04-01 11:33:59,236 mask_to_names() GetKeyState(VK_NUMLOCK)=1, names=['mod2'] 2020-04-01 11:33:59,240 mask_to_names() GetKeyState(VK_NUMLOCK)=1, names=['mod2'] 2020-04-01 11:34:02,953 mask_to_names() GetKeyState(VK_NUMLOCK)=1, names=['mod2'] 2020-04-01 11:34:02,959 mask_to_names() GetKeyState(VK_NUMLOCK)=1, names=['mod2'] On the server side : 2020-04-01 11:37:00,323 client 3 @22.200 mask_to_names() GetKeyState(VK_NUMLOCK)=1, names=['mod2'] 2020-04-01 11:37:00,329 client 3 @22.203 parse_key_event(, True)= 2020-04-01 11:37:00,388 client 3 @22.251 handle_key_action(ClientWindow(16), ) wid=16 2020-04-01 11:37:00,398 client 3 @22.253 key_handled_as_shortcut: shortcut(KP_5)=None 2020-04-01 11:37:00,398 client 3 @22.255 send_delayed_key() delayed_event=None 2020-04-01 11:37:00,412 client 3 @22.284 send_key_action(16, ) 2020-04-01 11:37:00,436 client 3 @22.290 mask_to_names() GetKeyState(VK_NUMLOCK)=1, names=['mod2'] 2020-04-01 11:37:00,436 will try levels: [0, 4, 1, 5, 2, 6, 3, 7] 2020-04-01 11:37:00,436 do_get_keycode(101, 'KP_5', True, ['mod2'], 0)=84 (level=1, shift=False, mode=0) 2020-04-01 11:37:00,436 adding 'shift' to modifiers 2020-04-01 11:37:00,436 process_key_action(['key-action', 16, b'KP_5', True, (b'mod2',), 65461, b'5', 101, 0]) server keycode=84, group=0 2020-04-01 11:37:00,436 set_keyboard_layout_group(0) config=KeyboardConfig(us / / None), current keyboard group=0 2020-04-01 11:37:00,436 filtered_modifiers_set(['mod2'])={'mod2'} 2020-04-01 11:37:00,437 filtered_modifiers_set(['mod2', 'shift'])={'mod2', 'shift'} 2020-04-01 11:37:00,437 make_keymask_match(['mod2', 'shift']) current mask: {'mod2'}, wanted: {'mod2', 'shift'}, ignoring=84/['KP_5'], keys_pressed={} 2020-04-01 11:37:00,437 change_mask(set(), False, remove) failed=[] 2020-04-01 11:37:00,437 keynames(shift)={'Shift_L', 'Shift_R'}, keycodes=[50, 62], nuisance=False, nuisance keys={'lock', 'mod2'} 2020-04-01 11:37:00,437 change_mask(add) ['mod2', 'shift'] modifier 'shift' using keycode 50 2020-04-01 11:37:00,437 change_mask({'shift'}, True, add) failed=[] 2020-04-01 11:37:00,438 is_modifier(84) not found 2020-04-01 11:37:00,438 handle_key((16, True, 'KP_5', 65461, 84, ['mod2', 'shift'], False, True)) 2020-04-01 11:37:00,438 handle keycode pressing 84: key 'KP_5' 2020-04-01 11:37:00,438 fake_key(84, True) 2020-04-01 11:37:00,438 scheduling key repeat timer with delay 500 for KP_5 / 84 2020-04-01 11:37:00,449 client 3 @22.316 parse_key_event(, False)= 2020-04-01 11:37:00,449 client 3 @22.318 handle_key_action(ClientWindow(16), ) wid=16 2020-04-01 11:37:00,449 client 3 @22.319 key_handled_as_shortcut: shortcut(KP_5)=None 2020-04-01 11:37:00,449 client 3 @22.319 send_delayed_key() delayed_event=None 2020-04-01 11:37:00,450 client 3 @22.320 send_key_action(16, ) 2020-04-01 11:37:00,450 process_key_action(['key-action', 16, b'KP_5', False, (b'mod2',), 65461, b'5', 101, 0]) server keycode=84, group=0 2020-04-01 11:37:00,450 set_keyboard_layout_group(0) config=KeyboardConfig(us / / None), current keyboard group=0 2020-04-01 11:37:00,450 filtered_modifiers_set(['shift', 'mod2'])={'mod2', 'shift'} 2020-04-01 11:37:00,450 filtered_modifiers_set(['mod2'])={'mod2'} 2020-04-01 11:37:00,450 make_keymask_match(['mod2']) current mask: {'mod2', 'shift'}, wanted: {'mod2'}, ignoring=84/['KP_5'], keys_pressed={84: 'KP_5'} 2020-04-01 11:37:00,451 keynames(shift)={'Shift_L', 'Shift_R'}, keycodes=[50, 62], nuisance=False, nuisance keys={'lock', 'mod2'} 2020-04-01 11:37:00,451 change_mask(remove) ['mod2'] modifier 'shift' using keycode 50 2020-04-01 11:37:00,451 change_mask({'shift'}, False, remove) failed=[] 2020-04-01 11:37:00,451 change_mask(set(), True, add) failed=[] 2020-04-01 11:37:00,452 is_modifier(84) not found 2020-04-01 11:37:00,452 handle_key((16, False, 'KP_5', 65461, 84, ['mod2'], False, True)) 2020-04-01 11:37:00,452 handle keycode unpressing 84: key 'KP_5' 2020-04-01 11:37:00,452 fake_key(84, False) 2020-04-01 11:37:08,817 client 3 @30.673 mask_to_names() GetKeyState(VK_NUMLOCK)=1, names=['mod2'] 2020-04-01 11:37:08,817 client 3 @30.679 mask_to_names() GetKeyState(VK_NUMLOCK)=1, names=['mod2'] 2020-04-01 11:37:08,817 clear_keys_pressed() Not sure what to look for! Thanks and Regards cl From antoine at nagafix.co.uk Wed Apr 1 16:58:48 2020 From: antoine at nagafix.co.uk (Antoine Martin) Date: Wed, 1 Apr 2020 22:58:48 +0700 Subject: [winswitch] Keypads number output wring value. In-Reply-To: References: Message-ID: On 01/04/2020 22:39, carl langlois via shifter-users wrote: > Hi, > > I am having issue with the keypad on my keyboard. When pressing number the > character display is not the number . i.e when i press 5 i got a E in the > terminal. Please create a ticket for this issue, and make sure to record as many details as possible from the bug reporting guidelines: https://xpra.org/trac/wiki/ReportingBugs The bare minimum would be the full OS and versions details for both client and server. And ideally more (logs, keyboard layout, etc): https://xpra.org/trac/wiki/Keyboard#ReportingBugs If this is a recent regression, you may be able to fix things by starting your server with: XPRA_SIMULATE_MODIFIERS=0 xpra start ... Cheers, Antoine > > I have enable the debug keyboard logging. When i press 5 on the keypad the > verbose output this. > > On the client side : > > 2020-04-01 11:33:58,596 parse_key_event( 0x0000000000bab400 (void at 0x0000000023108500)>, True)= object, contents: {'modifiers': ['mod2'], 'keyname': 'KP_5', 'keyval': > 65461, 'keycode': 101, 'group': 0, 'string': '5', 'pressed': True}> > 2020-04-01 11:33:58,639 handle_key_action(ClientWindow(16), object, contents: {'modifiers': ['mod2'], 'keyname': 'KP_5', 'keyval': > 65461, 'keycode': 101, 'group': 0, 'string': '5', 'pressed': True}>) wid=16 > 2020-04-01 11:33:58,670 key_handled_as_shortcut: shortcut(KP_5)=None > 2020-04-01 11:33:58,672 send_delayed_key() delayed_event=None > 2020-04-01 11:33:58,674 send_key_action(16, {'modifiers': ['mod2'], 'keyname': 'KP_5', 'keyval': 65461, 'keycode': 101, > 'group': 0, 'string': '5', 'pressed': True}>) > 2020-04-01 11:33:58,706 mask_to_names() > GetKeyState(VK_NUMLOCK)=1, names=['mod2'] > 2020-04-01 11:33:58,708 parse_key_event( 0x0000000000bab400 (void at 0x0000000023110460)>, False)= object, contents: {'modifiers': ['mod2'], 'keyname': 'KP_5', 'keyval': > 65461, 'keycode': 101, 'group': 0, 'string': '5', 'pressed': False}> > 2020-04-01 11:33:58,735 handle_key_action(ClientWindow(16), object, contents: {'modifiers': ['mod2'], 'keyname': 'KP_5', 'keyval': > 65461, 'keycode': 101, 'group': 0, 'string': '5', 'pressed': False}>) wid=16 > 2020-04-01 11:33:58,739 key_handled_as_shortcut: shortcut(KP_5)=None > 2020-04-01 11:33:58,741 send_delayed_key() delayed_event=None > 2020-04-01 11:33:58,774 send_key_action(16, {'modifiers': ['mod2'], 'keyname': 'KP_5', 'keyval': 65461, 'keycode': 101, > 'group': 0, 'string': '5', 'pressed': False}>) > 2020-04-01 11:33:59,201 mask_to_names() > GetKeyState(VK_NUMLOCK)=1, names=['mod2'] > 2020-04-01 11:33:59,208 mask_to_names() > GetKeyState(VK_NUMLOCK)=1, names=['mod2'] > 2020-04-01 11:33:59,236 mask_to_names() > GetKeyState(VK_NUMLOCK)=1, names=['mod2'] > 2020-04-01 11:33:59,240 mask_to_names() > GetKeyState(VK_NUMLOCK)=1, names=['mod2'] > 2020-04-01 11:34:02,953 mask_to_names() > GetKeyState(VK_NUMLOCK)=1, names=['mod2'] > 2020-04-01 11:34:02,959 mask_to_names( Gdk.ModifierType>) GetKeyState(VK_NUMLOCK)=1, names=['mod2'] > > > On the server side : > 2020-04-01 11:37:00,323 client 3 @22.200 mask_to_names( Gdk.ModifierType>) GetKeyState(VK_NUMLOCK)=1, names=['mod2'] > 2020-04-01 11:37:00,329 client 3 @22.203 parse_key_event( object at 0x0000000000bbd720 (void at 0x0000000023105a00)>, > True)= 'KP_5', 'keyval': 65461, 'keycode': 101, 'group': 0, 'string': '5', > 'pressed': True}> > 2020-04-01 11:37:00,388 client 3 @22.251 > handle_key_action(ClientWindow(16), {'modifiers': ['mod2'], 'keyname': 'KP_5', 'keyval': 65461, 'keycode': 101, > 'group': 0, 'string': '5', 'pressed': True}>) wid=16 > 2020-04-01 11:37:00,398 client 3 @22.253 key_handled_as_shortcut: > shortcut(KP_5)=None > 2020-04-01 11:37:00,398 client 3 @22.255 send_delayed_key() > delayed_event=None > 2020-04-01 11:37:00,412 client 3 @22.284 send_key_action(16, object, contents: {'modifiers': ['mod2'], 'keyname': 'KP_5', 'keyval': > 65461, 'keycode': 101, 'group': 0, 'string': '5', 'pressed': True}>) > 2020-04-01 11:37:00,436 client 3 @22.290 mask_to_names( Gdk.ModifierType>) GetKeyState(VK_NUMLOCK)=1, names=['mod2'] > 2020-04-01 11:37:00,436 will try levels: [0, 4, 1, 5, 2, 6, 3, 7] > 2020-04-01 11:37:00,436 do_get_keycode(101, 'KP_5', True, ['mod2'], 0)=84 > (level=1, shift=False, mode=0) > 2020-04-01 11:37:00,436 adding 'shift' to modifiers > 2020-04-01 11:37:00,436 process_key_action(['key-action', 16, b'KP_5', > True, (b'mod2',), 65461, b'5', 101, 0]) server keycode=84, group=0 > 2020-04-01 11:37:00,436 set_keyboard_layout_group(0) > config=KeyboardConfig(us / / None), current keyboard group=0 > 2020-04-01 11:37:00,436 filtered_modifiers_set(['mod2'])={'mod2'} > 2020-04-01 11:37:00,437 filtered_modifiers_set(['mod2', 'shift'])={'mod2', > 'shift'} > 2020-04-01 11:37:00,437 make_keymask_match(['mod2', 'shift']) current mask: > {'mod2'}, wanted: {'mod2', 'shift'}, ignoring=84/['KP_5'], keys_pressed={} > 2020-04-01 11:37:00,437 change_mask(set(), False, remove) failed=[] > 2020-04-01 11:37:00,437 keynames(shift)={'Shift_L', 'Shift_R'}, > keycodes=[50, 62], nuisance=False, nuisance keys={'lock', 'mod2'} > 2020-04-01 11:37:00,437 change_mask(add) ['mod2', 'shift'] modifier 'shift' > using keycode 50 > 2020-04-01 11:37:00,437 change_mask({'shift'}, True, add) failed=[] > 2020-04-01 11:37:00,438 is_modifier(84) not found > 2020-04-01 11:37:00,438 handle_key((16, True, 'KP_5', 65461, 84, ['mod2', > 'shift'], False, True)) > 2020-04-01 11:37:00,438 handle keycode pressing 84: key 'KP_5' > 2020-04-01 11:37:00,438 fake_key(84, True) > 2020-04-01 11:37:00,438 scheduling key repeat timer with delay 500 for KP_5 > / 84 > 2020-04-01 11:37:00,449 client 3 @22.316 parse_key_event( object at 0x0000000000bbd680 (void at 0x0000000023105a00)>, > False)= 'KP_5', 'keyval': 65461, 'keycode': 101, 'group': 0, 'string': '5', > 'pressed': False}> > 2020-04-01 11:37:00,449 client 3 @22.318 > handle_key_action(ClientWindow(16), {'modifiers': ['mod2'], 'keyname': 'KP_5', 'keyval': 65461, 'keycode': 101, > 'group': 0, 'string': '5', 'pressed': False}>) wid=16 > 2020-04-01 11:37:00,449 client 3 @22.319 key_handled_as_shortcut: > shortcut(KP_5)=None > 2020-04-01 11:37:00,449 client 3 @22.319 send_delayed_key() > delayed_event=None > 2020-04-01 11:37:00,450 client 3 @22.320 send_key_action(16, object, contents: {'modifiers': ['mod2'], 'keyname': 'KP_5', 'keyval': > 65461, 'keycode': 101, 'group': 0, 'string': '5', 'pressed': False}>) > 2020-04-01 11:37:00,450 process_key_action(['key-action', 16, b'KP_5', > False, (b'mod2',), 65461, b'5', 101, 0]) server keycode=84, group=0 > 2020-04-01 11:37:00,450 set_keyboard_layout_group(0) > config=KeyboardConfig(us / / None), current keyboard group=0 > 2020-04-01 11:37:00,450 filtered_modifiers_set(['shift', 'mod2'])={'mod2', > 'shift'} > 2020-04-01 11:37:00,450 filtered_modifiers_set(['mod2'])={'mod2'} > 2020-04-01 11:37:00,450 make_keymask_match(['mod2']) current mask: {'mod2', > 'shift'}, wanted: {'mod2'}, ignoring=84/['KP_5'], keys_pressed={84: 'KP_5'} > 2020-04-01 11:37:00,451 keynames(shift)={'Shift_L', 'Shift_R'}, > keycodes=[50, 62], nuisance=False, nuisance keys={'lock', 'mod2'} > 2020-04-01 11:37:00,451 change_mask(remove) ['mod2'] modifier 'shift' using > keycode 50 > 2020-04-01 11:37:00,451 change_mask({'shift'}, False, remove) failed=[] > 2020-04-01 11:37:00,451 change_mask(set(), True, add) failed=[] > 2020-04-01 11:37:00,452 is_modifier(84) not found > 2020-04-01 11:37:00,452 handle_key((16, False, 'KP_5', 65461, 84, ['mod2'], > False, True)) > 2020-04-01 11:37:00,452 handle keycode unpressing 84: key 'KP_5' > 2020-04-01 11:37:00,452 fake_key(84, False) > 2020-04-01 11:37:08,817 client 3 @30.673 mask_to_names( Gdk.ModifierType>) GetKeyState(VK_NUMLOCK)=1, names=['mod2'] > 2020-04-01 11:37:08,817 client 3 @30.679 mask_to_names( GDK_BUTTON1_MASK of type Gdk.ModifierType>) GetKeyState(VK_NUMLOCK)=1, > names=['mod2'] > 2020-04-01 11:37:08,817 clear_keys_pressed() > > > Not sure what to look for! > > > Thanks and Regards > > cl > _______________________________________________ > shifter-users mailing list > shifter-users at lists.devloop.org.uk > https://lists.devloop.org.uk/mailman/listinfo/shifter-users > From antoine at nagafix.co.uk Wed Apr 1 17:18:20 2020 From: antoine at nagafix.co.uk (Antoine Martin) Date: Wed, 1 Apr 2020 23:18:20 +0700 Subject: [winswitch] [ANNOUNCE] Xpra 3.0.8: nothing major In-Reply-To: <59185b49-771a-93c8-e9f9-c0130e305d26@devloop.org.uk> References: <59185b49-771a-93c8-e9f9-c0130e305d26@devloop.org.uk> Message-ID: On 01/04/2020 22:38, Antoine Martin via shifter-users wrote: > Hi, > > I managed to mess up the very last patch to go in this 3.0.8 release, > which caused a regression for those connecting to 3.0.8 servers via ssh > (system ssh, xpra's builtin ssh server was not affected). In case you upgrade and you still hit the same connection error: "Error: failed to receive anything, not an xpra server?" when doing a remote start like "xpra start ssh://user at HOST/ --start=app" You may need to first remove the corrupted run-xpra script on the server (it will then be re-generated): rm $XDG_RUNTIME_DIR/xpra/run-xpra Or you can also just run "xpra start --start=app" once, directly from a shell on the server. Cheers, Antoine > A patch has been applied and I have re-issued the builds for all the > affected platforms. (that's all of them except MS Windows) > Sorry about the hassle. > > Cheers, > Antoine > > > > On 31/03/2020 18:40, Antoine Martin wrote: >> Hi, >> >> There's a fair number of bug fixes in this release, but none of them are >> critical. >> The clipboard and keyboard (AltGr) handling are the most important ones >> in there, the rest is all fairly minor stuff. (platforms, warnings, etc) >> >> Release notes: >> * handling of dpi command line switch (correctly this time?) >> * bug report window not getting focus on MacOS >> * spurious ssh key warnings with newer versions of paramiko >> * AltGr mode with non-X11 clients, layout-group changes >> * rare unexpected client exit on MS Windows >> * MS Windows clipboard: >> update failures >> convert CRLF line endings >> * Clipboard: >> fix selection not shown as active in menus (MS Windows and MacOS) >> fix spurious warnings when sharing a session >> fix clipboard reset with python2 builds >> selection translation for outbound data >> support client applications that don't use TARGETS (ie: Motif) >> reject invalid targets >> * 'xpra upgrade' wrongly updating non-xpra displays >> * logging error in client geometry debugging output >> * pulseaudio start command with Ubuntu Xenial >> * spurious de-iconifications >> * handling of server control commands with python3 clients >> * UDP backport bug sending control packets >> * vfb getting killed on upgrade failure >> * proxy server cleanup: force forwarders to terminate >> * session info errors during client exit >> * printer cleanup errors with invalid UTF8 printer names >> * transient-for popup window workaround >> * missing libyuv csc module on MacOS >> * unicode errors saving xpra runner shell script with python3 >> * try to prevent conflicts with Fedora's packages >> * better detection of Wayland environments >> * use python3 (if installed) by default on Ubuntu Xenial >> * don't use Xdummy on arm, too slow >> * don't show 'Download' button that we can't honour >> * show all pressed keys according to X11 server in 'xpra info' >> * try harder not to use video for tiny areas >> >> Cheers, >> Antoine >> > > _______________________________________________ > shifter-users mailing list > shifter-users at lists.devloop.org.uk > https://lists.devloop.org.uk/mailman/listinfo/shifter-users > From jblack at xes-inc.com Tue Apr 7 16:58:44 2020 From: jblack at xes-inc.com (Jake Black) Date: Tue, 7 Apr 2020 10:58:44 -0500 (CDT) Subject: [winswitch] Start-desktop black screen and Shadow crashes on Ubuntu 18.04 Message-ID: <188689088.252946.1586275124067.JavaMail.zimbra@xes-inc.com> Greetings, I am trying to get xpra to show a full desktop on Ubuntu 18.04.3 (both client and server). I can successfully launch applications and they are stable. However, when I try to use "start-desktop", xpra launches, but just shows a black screen and I cannot interact with it. However, I can launch a desktop with lxsession. But I would really like to have a gnome session. When I close the window, the following error is printed: /usr/lib/python3/dist-packages/xpra/client/gtk_base/gtk_client_base.py:150: Warning: g_object_ref: assertion 'G_IS_OBJECT (object)' failed gtk_main() >From what I have found this is a pretty generic error and doesn't point to one thing. When I try to connect to an existing xsession with "shadow", the remote screens open, but when I interact with it by clicking in the window, they just close, "connection lost" is printed, the physical xsession crashes, and the following error is printed in /run/user/[uid]/xpra/:1.log: Gdk-Message: 10:12:32.098: Xpra: Fatal IO error 11 (Resource temporarily unavailable) on X server :1. I have found the following mailing list message that sounds familiar [ http://lists.devloop.org.uk/pipermail/shifter-users/2018-October/002231.html | http://lists.devloop.org.uk/pipermail/shifter-users/2018-October/002231.html ] And points to the following ticket: [ https://xpra.org/trac/ticket/1981#comment:5 | https://xpra.org/trac/ticket/1981#comment:5 ] And here the issue is reported as a bug in Ubuntu packaging and that the issue was resolved by removing the "gtk3-nocsd" and "libgtk3-nocsd0" packages. I have confirmed that these packages do not exist on either my server or my client. Additionally I have tried disabling opengl and using native opengl and the result does not change. Thanks, Jake From antoine at nagafix.co.uk Tue Apr 7 18:04:09 2020 From: antoine at nagafix.co.uk (Antoine Martin) Date: Wed, 8 Apr 2020 00:04:09 +0700 Subject: [winswitch] Start-desktop black screen and Shadow crashes on Ubuntu 18.04 In-Reply-To: <188689088.252946.1586275124067.JavaMail.zimbra@xes-inc.com> References: <188689088.252946.1586275124067.JavaMail.zimbra@xes-inc.com> Message-ID: <9d1e2d87-5d14-8125-27c1-0a3c8f9a6bd4@nagafix.co.uk> On 07/04/2020 22:58, Jake Black via shifter-users wrote: > Greetings, > > I am trying to get xpra to show a full desktop on Ubuntu 18.04.3 (both client and server). I can successfully launch applications and they are stable. However, when I try to use "start-desktop", xpra launches, but just shows a black screen and I cannot interact with it. You need to start a window manager. ie: xpra start-desktop \ --start-child=fluxbox --exit-with-children \ --attach=yes Or an xterm and you can try starting your window manager from there. > However, I can launch a desktop with lxsession. But I would really like to have a gnome session. On Fedora 31, using --start=gnome-session, I get the dreaded "oh no, something went wrong dialog" but after dismissing it the session is actually perfectly usable. On Ubuntu 18.04 and 19.10, it just does not work at all. Perhaps GNOME requires OpenGL acceleration to work properly? (since I couldn't find a virtualgl package for Ubuntu, I didn't try this potential workaround) Generally speaking, when displaying things remotely via xpra, you're much better off running a lightweight desktop like fluxbox than a heavy one like GNOME or KDE. > When I close the window, the following error is printed: > > /usr/lib/python3/dist-packages/xpra/client/gtk_base/gtk_client_base.py:150: Warning: g_object_ref: assertion 'G_IS_OBJECT (object)' failed > gtk_main() This is completely unrelated and you can safely ignore this message. > From what I have found this is a pretty generic error and doesn't point to one thing. Correct. > When I try to connect to an existing xsession with "shadow", the remote screens open, but when I interact with it by clicking in the window, they just close, "connection lost" is printed, the physical xsession crashes, Do you mean the xpra server? (not the local X11 session) > and the following error is printed in /run/user/[uid]/xpra/:1.log: > > Gdk-Message: 10:12:32.098: Xpra: Fatal IO error 11 (Resource temporarily unavailable) on X server :1. That's usually a sign that the X11 server has crashed. I've just quickly tried shadowing 18.04 and 19.10 test systems, and I am not seeing any crashes or other obvious bugs. I suspect that there must be something different with your system. Please make sure that you're running the latest official version. > I have found the following mailing list message that sounds familiar > [ http://lists.devloop.org.uk/pipermail/shifter-users/2018-October/002231.html | http://lists.devloop.org.uk/pipermail/shifter-users/2018-October/002231.html ] > > And points to the following ticket: > > [ https://xpra.org/trac/ticket/1981#comment:5 | https://xpra.org/trac/ticket/1981#comment:5 ] > > And here the issue is reported as a bug in Ubuntu packaging and that the issue was resolved by removing the "gtk3-nocsd" and "libgtk3-nocsd0" packages. I have confirmed that these packages do not exist on either my server or my client. That's unlikely to be related if you don't have gtk3-nocsd installed. Please post the full details of your error, preferably in a new ticket. > > Additionally I have tried disabling opengl and using native opengl and the result does not change. Cheers, Antoine > > > Thanks, > > Jake > _______________________________________________ > shifter-users mailing list > shifter-users at lists.devloop.org.uk > https://lists.devloop.org.uk/mailman/listinfo/shifter-users > From timlee126 at yahoo.com Wed Apr 8 00:57:28 2020 From: timlee126 at yahoo.com (Tim) Date: Tue, 7 Apr 2020 23:57:28 +0000 (UTC) Subject: [winswitch] Fail to connect to a XPRA server References: <2135199104.2230916.1586303848962.ref@mail.yahoo.com> Message-ID: <2135199104.2230916.1586303848962@mail.yahoo.com> Why can't I connect to the remote xpra server? ??? t at ocean:/u/l/jvm$ xpra shadow ssh:"onion":0 ?? ? ??? (xpra:24602): dbind-WARNING **: 21:28:35.326: Couldn't register with accessibility bus: Did not receive a reply. Possible causes include: the remote application did not send a reply, the message bus security policy blocked the reply, the reply timeout expired, or the network connection was broken. ??? 2020-04-06 21:28:35,442 Xpra GTK3 X11 client version 3.0.7-r25627 64-bit ??? 2020-04-06 21:28:35,721? running on Linux Ubuntu 18.04 bionic ??? 2020-04-06 21:28:35,724? window manager is 'Openbox' ??? 2020-04-06 21:28:35,770 Warning: failed to import opencv: ??? 2020-04-06 21:28:35,770? No module named 'cv2' ??? 2020-04-06 21:28:35,770? webcam forwarding is disabled ?? ? ??? (Xpra-Audio-query:24619): dbind-WARNING **: 21:28:36.798: Couldn't register with accessibility bus: Did not receive a reply. Possible causes include: the remote application did not send a reply, the message bus security policy blocked the reply, the reply timeout expired, or the network connection was broken. ??? 2020-04-06 21:28:36,926 GStreamer version 1.14.5 for Python 3.6.9 64-bit ??? 2020-04-06 21:28:37,314 No OpenGL_accelerate module loaded: No module named 'OpenGL_accelerate' ??? 2020-04-06 21:28:37,772 Error: cannot handle window transparency ??? 2020-04-06 21:28:37,772? screen is not composited ??? 2020-04-06 21:28:37,848 Warning: vendor 'Intel Open Source Technology Center' is greylisted, ??? 2020-04-06 21:28:37,848? you may want to turn off OpenGL if you encounter bugs ??? 2020-04-06 21:28:38,043 Error: cannot handle window transparency ??? 2020-04-06 21:28:38,043? screen is not composited ??? 2020-04-06 21:28:38,053 Warning: window 4294967295 changed its transparency attribute ??? 2020-04-06 21:28:38,054? from False to True, behaviour is undefined ??? 2020-04-06 21:28:38,133 OpenGL enabled with Mesa DRI Mobile Intel? GM45 Express Chipset ??? 2020-04-06 21:28:38,407 Connected (version 2.0, client OpenSSH_7.6p1) ??? 2020-04-06 21:28:38,961 loaded RSA private key from '/home/t/.ssh/id_rsa' ??? 2020-04-06 21:28:38,981 Authentication (publickey) successful! ??? 2020-04-06 21:28:39,275? keyboard settings: rules=evdev, model=pc105, layout=us ??? 2020-04-06 21:28:39,356? desktop size is 1440x900 with 1 screen: ??? 2020-04-06 21:28:39,356?? :0.0 (381x238 mm - DPI: 96x96) workarea: 1440x876 ??? 2020-04-06 21:28:39,357???? LVDS-1 (303x190 mm - DPI: 120x120) ??? 2020-04-06 21:28:39,432 no ethtool interface speed available for wlx801f02b5c389 ??? ^C ??? 2020-04-06 21:28:55,157 got signal SIGINT ??? 2020-04-06 21:28:55,157 exiting My local xpra client is ??? t at ocean:/u/l/jvm$ xpra --version ??? xpra v3.0.7-r25627 ?? ? A remote xpra (server, or client or both?) is ??? t at ocean:/u/l/jvm$ ssh t at onion ??? t at onion:~$ xpra --version ??? xpra v3.0.8-r25889 ??? t at onion:~$ logout ??? Connection to onion closed. Both local and remote are running Lubuntu 18.04. ? ? Thanks. From timlee126 at yahoo.com Wed Apr 8 02:27:27 2020 From: timlee126 at yahoo.com (Tim) Date: Wed, 8 Apr 2020 01:27:27 +0000 (UTC) Subject: [winswitch] xpra client window on 14" laptop doesn't show full screen of 15" laptop running xpra server References: <1183712019.2251157.1586309247941.ref@mail.yahoo.com> Message-ID: <1183712019.2251157.1586309247941@mail.yahoo.com> The xpra client window?? on a 14" laptop shows the screen of the 15" laptop that runs the xpra server only partially (a part near the bottom and a part near the right are cut off). The screenshot below shows the screen of the 14" laptop with the xpra client window inside: https://i.stack.imgur.com/GA9rX.png Both laptops are running Lubuntu 18.04. The remote xpra version and screen dimension and resolution are: ??? $ xpra --version ??? xpra v3.0.8-r25889 ??? $ xdpyinfo | grep dimensions ????? dimensions:??? 1920x1080 pixels (508x285 millimeters) The local xpra version (shown different from the remote one, but might be incorrect, see https://unix.stackexchange.com/questions/578622/have-i-upgraded-a-package-and-why-is-there-a-mismatch-of-its-version-shown) and screen dimension and resolution are: ??? $ xpra --version ??? xpra v3.0.7-r25627 ?? ? ??? $ xdpyinfo | grep dimensions ????? dimensions:??? 1440x900 pixels (381x238 millimeters) Here is the command that I ran to connect ??? $ xpra shadow ssh:"t at onion":0 ?? ? ??? (xpra:16983): dbind-WARNING **: 20:13:10.163: Couldn't register with accessibility bus: Did not receive a reply. Possible causes include: the remote application did not send a reply, the message bus security policy blocked the reply, the reply timeout expired, or the network connection was broken. ??? 2020-04-07 20:13:10,318 Xpra GTK3 X11 client version 3.0.7-r25627 64-bit ??? 2020-04-07 20:13:10,613? running on Linux Ubuntu 18.04 bionic ??? 2020-04-07 20:13:10,615? window manager is 'Openbox' ??? 2020-04-07 20:13:10,697 Warning: failed to import opencv: ??? 2020-04-07 20:13:10,697? No module named 'cv2' ??? 2020-04-07 20:13:10,698? webcam forwarding is disabled ?? ? ??? (Xpra-Audio-query:17000): dbind-WARNING **: 20:13:11.757: Couldn't register with accessibility bus: Did not receive a reply. Possible causes include: the remote application did not send a reply, the message bus security policy blocked the reply, the reply timeout expired, or the network connection was broken. ??? 2020-04-07 20:13:11,895 GStreamer version 1.14.5 for Python 3.6.9 64-bit ??? 2020-04-07 20:13:12,374 No OpenGL_accelerate module loaded: No module named 'OpenGL_accelerate' ??? 2020-04-07 20:13:13,002 Error: cannot handle window transparency ??? 2020-04-07 20:13:13,002? screen is not composited ??? 2020-04-07 20:13:13,086 Warning: vendor 'Intel Open Source Technology Center' is greylisted, ??? 2020-04-07 20:13:13,086? you may want to turn off OpenGL if you encounter bugs ??? 2020-04-07 20:13:13,349 Error: cannot handle window transparency ??? 2020-04-07 20:13:13,349? screen is not composited ??? 2020-04-07 20:13:13,360 Warning: window 4294967295 changed its transparency attribute ??? 2020-04-07 20:13:13,361? from False to True, behaviour is undefined ??? 2020-04-07 20:13:13,450 OpenGL enabled with Mesa DRI Mobile Intel? GM45 Express Chipset ??? 2020-04-07 20:13:13,730 Connected (version 2.0, client OpenSSH_7.6p1) ??? 2020-04-07 20:13:14,386 loaded RSA private key from '/home/t/.ssh/id_rsa' ??? 2020-04-07 20:13:14,405 Authentication (publickey) successful! ??? 2020-04-07 20:13:14,736? keyboard settings: rules=evdev, model=pc105, layout=us ??? 2020-04-07 20:13:14,834? desktop size is 1440x900 with 1 screen: ??? 2020-04-07 20:13:14,835?? :0.0 (381x238 mm - DPI: 96x96) workarea: 1440x876 ??? 2020-04-07 20:13:14,835???? LVDS-1 (303x190 mm - DPI: 120x120) ??? 2020-04-07 20:13:14,909 no ethtool interface speed available for wlx801f02b5c389 ??? 2020-04-07 20:13:20,742 enabled remote logging ??? 2020-04-07 20:13:20,744 Xpra GTK3 shadow server version 3.0.8-r25889 64-bit ??? 2020-04-07 20:13:20,746? running on Linux Ubuntu 18.04 bionic ??? 2020-04-07 20:13:21,120 server does not support xi input devices ??? 2020-04-07 20:13:21,123? server uses: auto ?? ? ??? (Xpra-Audio-play:17029): dbind-WARNING **: 20:13:22.207: Couldn't register with accessibility bus: Did not receive a reply. Possible causes include: the remote application did not send a reply, the message bus security policy blocked the reply, the reply timeout expired, or the network connection was broken. ??? 2020-04-07 20:13:22,299 sound output using pulseaudio device: ??? 2020-04-07 20:13:22,300 sound output? 'Built-in Audio Analog Stereo' ??? 2020-04-07 20:13:22,898 sound output using 'opus' audio codec From antoine at nagafix.co.uk Wed Apr 8 07:06:07 2020 From: antoine at nagafix.co.uk (Antoine Martin) Date: Wed, 8 Apr 2020 13:06:07 +0700 Subject: [winswitch] Fail to connect to a XPRA server In-Reply-To: <2135199104.2230916.1586303848962@mail.yahoo.com> References: <2135199104.2230916.1586303848962.ref@mail.yahoo.com> <2135199104.2230916.1586303848962@mail.yahoo.com> Message-ID: On 08/04/2020 06:57, Tim via shifter-users wrote: > Why can't I connect to the remote xpra server? > > ??? t at ocean:/u/l/jvm$ xpra shadow ssh:"onion":0 (..) > ??? 2020-04-06 21:28:38,981 Authentication (publickey) successful! > ??? 2020-04-06 21:28:39,275? keyboard settings: rules=evdev, model=pc105, layout=us > ??? 2020-04-06 21:28:39,356? desktop size is 1440x900 with 1 screen: > ??? 2020-04-06 21:28:39,356?? :0.0 (381x238 mm - DPI: 96x96) workarea: 1440x876 > ??? 2020-04-06 21:28:39,357???? LVDS-1 (303x190 mm - DPI: 120x120) > ??? 2020-04-06 21:28:39,432 no ethtool interface speed available for wlx801f02b5c389 > ??? ^C > ??? 2020-04-06 21:28:55,157 got signal SIGINT > ??? 2020-04-06 21:28:55,157 exiting There's nothing obvious here. Try starting the shadow server manually from an ssh session: xpra shadow :0 and keep an eye on its log file, then attach from the client: xpra attach ssh://onion/0 Cheers, Antoine > My local xpra client is > > ??? t at ocean:/u/l/jvm$ xpra --version > ??? xpra v3.0.7-r25627 > ?? ? > A remote xpra (server, or client or both?) is > > ??? t at ocean:/u/l/jvm$ ssh t at onion > ??? t at onion:~$ xpra --version > ??? xpra v3.0.8-r25889 > ??? t at onion:~$ logout > ??? Connection to onion closed. > > Both local and remote are running Lubuntu 18.04. > ? ? > Thanks. From antoine at nagafix.co.uk Wed Apr 8 07:44:33 2020 From: antoine at nagafix.co.uk (Antoine Martin) Date: Wed, 8 Apr 2020 13:44:33 +0700 Subject: [winswitch] xpra client window on 14" laptop doesn't show full screen of 15" laptop running xpra server In-Reply-To: <1183712019.2251157.1586309247941@mail.yahoo.com> References: <1183712019.2251157.1586309247941.ref@mail.yahoo.com> <1183712019.2251157.1586309247941@mail.yahoo.com> Message-ID: On 08/04/2020 08:27, Tim via shifter-users wrote: > The xpra client window?? on a 14" laptop shows the screen of the 15" laptop that runs the xpra server only partially (a part near the bottom and a part near the right are cut off). You have two options: * use --desktop-scaling=0.8 to downscale the shadow window, you can also access this scaling setting from xpra's system tray menu * try --desktop-fullscreen=yes (your mileage may vary) In the future, we will also have the option of using scrollbars for windows that don't fit. > The screenshot below shows the screen of the 14" laptop with the xpra client window inside: https://i.stack.imgur.com/GA9rX.png > Both laptops are running Lubuntu 18.04. So you did resolve your shadow problems from your earlier email? How? > The remote xpra version and screen dimension and resolution are: > > ??? $ xpra --version > ??? xpra v3.0.8-r25889 > > ??? $ xdpyinfo | grep dimensions > ????? dimensions:??? 1920x1080 pixels (508x285 millimeters) > > The local xpra version (shown different from the remote one, but might be incorrect, see https://unix.stackexchange.com/questions/578622/have-i-upgraded-a-package-and-why-is-there-a-mismatch-of-its-version-shown) and screen dimension and resolution are: > > ??? $ xpra --version > ??? xpra v3.0.7-r25627 That's not normal, something is wrong with your installation. > ?? ? > ??? $ xdpyinfo | grep dimensions > ????? dimensions:??? 1440x900 pixels (381x238 millimeters) (..) Cheers, Antoine From stdedos at gmail.com Sat Apr 11 19:12:39 2020 From: stdedos at gmail.com (=?UTF-8?B?zqPPhM6xz43Pgc6/z4Igzp3PhM6tzr3PhM6/z4I=?=) Date: Sat, 11 Apr 2020 21:12:39 +0300 Subject: [winswitch] Discussion about https://xpra.org/trac/changeset/26090/xpra Message-ID: Hello Antoine, Coffee-break discussion: I understand why you would maintain v2.[45], v3.x (thank you :-)) ... Why on earth maintain v1.x? Is it some "smaller" version, supporting ages old systems ... what? With regards, Ntentos Stavros From antoine at nagafix.co.uk Sun Apr 12 07:34:11 2020 From: antoine at nagafix.co.uk (Antoine Martin) Date: Sun, 12 Apr 2020 13:34:11 +0700 Subject: [winswitch] Discussion about https://xpra.org/trac/changeset/26090/xpra In-Reply-To: References: Message-ID: <37ed0307-52c5-7356-8f61-5a68f0da49f1@nagafix.co.uk> On 12/04/2020 01:12, ??????? ??????? via shifter-users wrote: > Hello Antoine, > > Coffee-break discussion: I understand why you would maintain v2.[45], v3.x > (thank you :-)) ... V2.x are not LTS versions but v1.x is. > Why on earth maintain v1.x? Is it some "smaller" version, supporting ages > old systems ... what? xpra v1.x was the last branch to support Python versions 2.6 and older, which includes OSes like RHEL / CentOS 6.x Like it or not, RHEL 6.x is going to be supported by RedHat until 2024 under the "Extended Lifecycle Support" schedule. Preserving basic backwards compatibility is not that much of a difficulty anyway. That said, it is time to move on, and the next 1.x stable update is likely to be the last. Cheers, Antoine > > With regards, > Ntentos Stavros > _______________________________________________ > shifter-users mailing list > shifter-users at lists.devloop.org.uk > https://lists.devloop.org.uk/mailman/listinfo/shifter-users > From antoine at devloop.org.uk Tue Apr 14 18:11:42 2020 From: antoine at devloop.org.uk (Antoine Martin) Date: Wed, 15 Apr 2020 00:11:42 +0700 Subject: [winswitch] [ANNOUNCE] Xpra 3.0.9: critical fixes Message-ID: Hi, There's a large number of bug fixes in this release, and quite a few of these bugs are critical. In particular: * the HTML5 XSS is trivial to exploit * the focus issues can be a real nuisance (mostly affects GTK3) * then there are bugs that could cause the server to crash - never a good thing, or render it unbearably slow * the MS Windows GTK3 client could get thoroughly confused by some window constraints (ie: Firefox maximized) Then there's the usual set of keyboard, clipboard, system tray, NVENC, platform and application quirks, etc Also noteworthy is the update to the application hints list: http://xpra.org/trac/changeset/25902/xpra If your preferred applications are not in that list, it is a good idea to submit them: this helps the xpra server make the correct decisions. Updating is strongly recommended. Release notes: * fix HTML5 client XSS * fix focus issues * fix server hangs and unpainted windows due to invalid dimensions * fix window icons and cursors visual corruption or crashes * fix unbearably slow screen updates when vrefresh value is invalid * fix client OpenGL crashes with remote starts (run probe first) * fix keypad regression (AltGr changes, again) * fix MS Windows shadow servers * fix race condition collecting latency data for session info * fix missing pycuda dependency for NVENC on MS Windows * fix missing signatures on MS Windows EXE and MSI builds * fix scoring of NVENC codec (should be used ahead of sofware encoders) * fix start menu updates getting lost * fix OpenGL diagnostic messages getting cut off * fix change of encoding from the system tray * fix quality and speed changes from the system tray * fix clipboard and popup menu problems with Eclipse * fix name of shadow server process on MS Windows, used in log files * fix audio query subprocess with MS shadow servers * fix unlikely race condition in speaker forwarding shutdown sequence * fix desktop-fullscreen=yes automatic scaling * fix missing bug report data due to uninitialized X11 bindings * fix bug report error if we cannot grab a screenshot (ie: Wayland) * fix error saving binary data with python3 builds (ie: bug report tool) * fix errors querying the display under Wayland * hide scary stacktraces when shutting down ssh proxy commands * if we have an ssh password we can use, try that first * blacklist VirtualBox's SVGA3D OpenGL driver (fixes black windows) * don't use webp with image dimensions beyond its limits * use regular X11 system tray with i3 * hide confusing named-pipe prefix in empty output on MS Windows * HTML5 fixes: invalid IMG tag, reserved keyword, language detection, draw debugging * don't try to read /proc if not on Posix * support more versions of Visual Studio when compiling CUDA kernels * use a more specific python3 interpreter version * update application content-type mappings * remove out of date information from man page * MacOS: openssl 1.1.1f Cheers, Antoine From crl.langlois at gmail.com Thu Apr 16 18:35:57 2020 From: crl.langlois at gmail.com (carl langlois) Date: Thu, 16 Apr 2020 13:35:57 -0400 Subject: [winswitch] Focus issue Message-ID: Hi, Using 3.0.9 on both server(Ubuntu 19.10) and client(windows 10) the mouse does not seem to click at the good location. Like if all focus is wrong between the windows and the mouse. Thanks and Regards Carl From antoine at nagafix.co.uk Fri Apr 17 07:55:36 2020 From: antoine at nagafix.co.uk (Antoine Martin) Date: Fri, 17 Apr 2020 13:55:36 +0700 Subject: [winswitch] Focus issue In-Reply-To: References: Message-ID: <120866b1-bf6a-55b0-6381-41c704ecddb3@nagafix.co.uk> On 17/04/2020 00:35, carl langlois via shifter-users wrote: > Hi, > > Using 3.0.9 on both server(Ubuntu 19.10) and client(windows 10) the mouse > does not seem to click at the good location. Like if all focus is wrong > between the windows and the mouse. Is this with all applications or just a specific one? For all its windows, or maybe just popup menus? Does it not work with a plain xterm, or maybe your window is maximized? If the offset large or small? Are you using Wayland client-side perhaps? There are a number of tickets related to pointer offsets, but they generally involve Java or proprietary applications. Here's a ticket which may be related: http://xpra.org/trac/ticket/2600 If this is the right place, please add your information there. Cheers, Antoine > > Thanks and Regards > > Carl > _______________________________________________ > shifter-users mailing list > shifter-users at lists.devloop.org.uk > https://lists.devloop.org.uk/mailman/listinfo/shifter-users > From crl.langlois at gmail.com Fri Apr 17 13:26:20 2020 From: crl.langlois at gmail.com (carl langlois) Date: Fri, 17 Apr 2020 08:26:20 -0400 Subject: [winswitch] Focus issue In-Reply-To: <120866b1-bf6a-55b0-6381-41c704ecddb3@nagafix.co.uk> References: <120866b1-bf6a-55b0-6381-41c704ecddb3@nagafix.co.uk> Message-ID: Hi, When it happen all the windows seems to be affected. It seem that all the. It look to me that some windows(application) grab the focus and all the other application does not received even. I can currently easily reproduce the situation where a windows freeze (i suspect a application bugs) . It seem that in this case the application enter a weird state and freeze all other application managed by Xpra. Even new application started from the tray menu are not responding to the mouse. if i ssh to the workstation and kill the application everything is start working again. I think this particular problem is coming from a bad behavior of Rabbitvcs.. I am not a expert in this kind of software but the fact that all the windows are frozen seems wrong. Not sure how to tackle this situation. Not related to java application and i have seen this behavior with other perl/tk application, dophin file manager etc. Carl. On Fri, Apr 17, 2020 at 2:55 AM Antoine Martin via shifter-users < shifter-users at lists.devloop.org.uk> wrote: > On 17/04/2020 00:35, carl langlois via shifter-users wrote: > > Hi, > > > > Using 3.0.9 on both server(Ubuntu 19.10) and client(windows 10) the mouse > > does not seem to click at the good location. Like if all focus is wrong > > between the windows and the mouse. > Is this with all applications or just a specific one? For all its > windows, or maybe just popup menus? > Does it not work with a plain xterm, or maybe your window is maximized? > If the offset large or small? Are you using Wayland client-side perhaps? > > There are a number of tickets related to pointer offsets, but they > generally involve Java or proprietary applications. Here's a ticket > which may be related: > http://xpra.org/trac/ticket/2600 > If this is the right place, please add your information there. > > Cheers, > Antoine > > > > > > Thanks and Regards > > > > Carl > > _______________________________________________ > > shifter-users mailing list > > shifter-users at lists.devloop.org.uk > > https://lists.devloop.org.uk/mailman/listinfo/shifter-users > > > > _______________________________________________ > shifter-users mailing list > shifter-users at lists.devloop.org.uk > https://lists.devloop.org.uk/mailman/listinfo/shifter-users > From antoine at nagafix.co.uk Fri Apr 17 16:16:08 2020 From: antoine at nagafix.co.uk (Antoine Martin) Date: Fri, 17 Apr 2020 22:16:08 +0700 Subject: [winswitch] Focus issue In-Reply-To: References: <120866b1-bf6a-55b0-6381-41c704ecddb3@nagafix.co.uk> Message-ID: <7373749a-1d4f-4624-a73b-d6a780d213b6@nagafix.co.uk> On 17/04/2020 19:26, carl langlois wrote: > Hi, > When it happen all the windows seems to be affected. It seem that all > the. It look to me that some windows(application) grab the focus and all > the other application does not received even. Please see --modal-windows=no A modal window will stay on top of all other windows unless you turn off this feature. > I can currently?easily?reproduce?the situation where a windows freeze (i > suspect a application bugs) . It seem that in this case the application > enter a weird state and freeze all other application managed by Xpra. > Even new application started from the tray menu are not responding to > the mouse. Unless the server eventually disconnects, this sounds like a modal window or a grab. > if i ssh to the workstation and kill the application everything is start > working again. I think this particular problem is coming from a bad > behavior of Rabbitvcs.. I am not a expert in this kind of software but > the fact that all the windows are frozen seems wrong. Could be, it could also be a bug in xpra. > Not sure how to tackle this situation. Not related to java application > and i have seen this behavior with other perl/tk application, > dophin?file manager etc. I've just tried it in an Ubuntu 19.10 virtual machine and I couldn't see any obvious problems. Do I have to use a specific dialog or action to trigger this bug? Assuming that this isn't a modal window problem, would you mind creating a ticket for this problem? Thanks, Antoine > Carl. > > > > > On Fri, Apr 17, 2020 at 2:55 AM Antoine Martin via shifter-users > > wrote: > > On 17/04/2020 00:35, carl langlois via shifter-users wrote: > > Hi, > > > > Using 3.0.9 on both server(Ubuntu 19.10) and client(windows 10) > the mouse > > does not seem to click at the good location. Like if all focus is > wrong > > between the windows and the mouse. > Is this with all applications or just a specific one? For all its > windows, or maybe just popup menus? > Does it not work with a plain xterm, or maybe your window is maximized? > If the offset large or small? Are you using Wayland client-side perhaps? > > There are a number of tickets related to pointer offsets, but they > generally involve Java or proprietary applications. Here's a ticket > which may be related: > http://xpra.org/trac/ticket/2600 > If this is the right place, please add your information there. > > Cheers, > Antoine > > > > > > Thanks and Regards > > > > Carl > > _______________________________________________ > > shifter-users mailing list > > shifter-users at lists.devloop.org.uk > > > https://lists.devloop.org.uk/mailman/listinfo/shifter-users > > > > _______________________________________________ > shifter-users mailing list > shifter-users at lists.devloop.org.uk > > https://lists.devloop.org.uk/mailman/listinfo/shifter-users > From antoine at nagafix.co.uk Fri Apr 17 17:12:35 2020 From: antoine at nagafix.co.uk (Antoine Martin) Date: Fri, 17 Apr 2020 23:12:35 +0700 Subject: [winswitch] Focus issue In-Reply-To: References: <120866b1-bf6a-55b0-6381-41c704ecddb3@nagafix.co.uk> <7373749a-1d4f-4624-a73b-d6a780d213b6@nagafix.co.uk> Message-ID: <1ced558c-b659-2dce-93ad-70fcac650e8c@nagafix.co.uk> On 17/04/2020 23:02, carl langlois wrote: > HI, > > You are right this seem to be a modal issue. Reproducing the problem > with modal disable fix the freezing issue. Great. FYI: starting with version 4.0, modal-windows will be disabled by default. > To trigger this i need to go in the check for modification > dialog->remote changes->(you need to have some changes) if you double > click on one of the changes it will popup two windows. On my side the > two popup are freezing(Rabbitvcs issue ..) Does it also freeze without xpra? > but with modal disable i can > xkill the rabbitvcs windows and resume normal operation. > > This morning?i had another phenomena like this one with other perl/tk > applications but i was able to recover because not all windows?was > affected.. if it happen again i will try disabling the modal windows. > > Great hints!! > Since it is a modal issue to you need me to open a ticket? No, it doesn't look like one is needed. (or maybe for the rabbitcvs freeze?) Cheers, Antoine > Carl?? > > > > On Fri, Apr 17, 2020 at 11:16 AM Antoine Martin > wrote: > > On 17/04/2020 19:26, carl langlois wrote: > > Hi, > > When it happen all the windows seems to be affected. It seem that all > > the. It look to me that some windows(application) grab the focus > and all > > the other application does not received even. > Please see --modal-windows=no > A modal window will stay on top of all other windows unless you turn off > this feature. > > I can currently?easily?reproduce?the situation where a windows > freeze (i > > suspect a application bugs) . It seem that in this case the > application > > enter a weird state and freeze all other application managed by Xpra. > > Even new application started from the tray menu are not responding to > > the mouse. > Unless the server eventually disconnects, this sounds like a modal > window or a grab. > > > if i ssh to the workstation and kill the application everything is > start > > working again. I think this particular problem is coming from a bad > > behavior of Rabbitvcs.. I am not a expert in this kind of software but > > the fact that all the windows are frozen seems wrong. > Could be, it could also be a bug in xpra. > > Not sure how to tackle this situation. Not related to java application > > and i have seen this behavior with other perl/tk application, > > dophin?file manager etc. > I've just tried it in an Ubuntu 19.10 virtual machine and I couldn't see > any obvious problems. > Do I have to use a specific dialog or action to trigger this bug? > > Assuming that this isn't a modal window problem, would you mind creating > a ticket for this problem? > > Thanks, > Antoine > > > > Carl. > > > > > > > > > > On Fri, Apr 17, 2020 at 2:55 AM Antoine Martin via shifter-users > > > > >> wrote: > > > >? ? ?On 17/04/2020 00:35, carl langlois via shifter-users wrote: > >? ? ?> Hi, > >? ? ?> > >? ? ?> Using 3.0.9 on both server(Ubuntu 19.10) and client(windows 10) > >? ? ?the mouse > >? ? ?> does not seem to click at the good location. Like if all > focus is > >? ? ?wrong > >? ? ?> between the windows and the mouse. > >? ? ?Is this with all applications or just a specific one? For all its > >? ? ?windows, or maybe just popup menus? > >? ? ?Does it not work with a plain xterm, or maybe your window is > maximized? > >? ? ?If the offset large or small? Are you using Wayland > client-side perhaps? > > > >? ? ?There are a number of tickets related to pointer offsets, but they > >? ? ?generally involve Java or proprietary applications. Here's a > ticket > >? ? ?which may be related: > >? ? ?http://xpra.org/trac/ticket/2600 > >? ? ?If this is the right place, please add your information there. > > > >? ? ?Cheers, > >? ? ?Antoine > > > > > >? ? ?> > >? ? ?> Thanks and Regards > >? ? ?> > >? ? ?> Carl > >? ? ?> _______________________________________________ > >? ? ?> shifter-users mailing list > >? ? ?> shifter-users at lists.devloop.org.uk > > >? ? ? > > >? ? ?> https://lists.devloop.org.uk/mailman/listinfo/shifter-users > >? ? ?> > > > >? ? ?_______________________________________________ > >? ? ?shifter-users mailing list > >? ? ?shifter-users at lists.devloop.org.uk > > >? ? ? > > >? ? ?https://lists.devloop.org.uk/mailman/listinfo/shifter-users > > > From crl.langlois at gmail.com Fri Apr 17 17:02:50 2020 From: crl.langlois at gmail.com (carl langlois) Date: Fri, 17 Apr 2020 12:02:50 -0400 Subject: [winswitch] Focus issue In-Reply-To: <7373749a-1d4f-4624-a73b-d6a780d213b6@nagafix.co.uk> References: <120866b1-bf6a-55b0-6381-41c704ecddb3@nagafix.co.uk> <7373749a-1d4f-4624-a73b-d6a780d213b6@nagafix.co.uk> Message-ID: HI, You are right this seem to be a modal issue. Reproducing the problem with modal disable fix the freezing issue. To trigger this i need to go in the check for modification dialog->remote changes->(you need to have some changes) if you double click on one of the changes it will popup two windows. On my side the two popup are freezing(Rabbitvcs issue ..) but with modal disable i can xkill the rabbitvcs windows and resume normal operation. This morning i had another phenomena like this one with other perl/tk applications but i was able to recover because not all windows was affected.. if it happen again i will try disabling the modal windows. Great hints!! Since it is a modal issue to you need me to open a ticket? Carl On Fri, Apr 17, 2020 at 11:16 AM Antoine Martin wrote: > On 17/04/2020 19:26, carl langlois wrote: > > Hi, > > When it happen all the windows seems to be affected. It seem that all > > the. It look to me that some windows(application) grab the focus and all > > the other application does not received even. > Please see --modal-windows=no > A modal window will stay on top of all other windows unless you turn off > this feature. > > I can currently easily reproduce the situation where a windows freeze (i > > suspect a application bugs) . It seem that in this case the application > > enter a weird state and freeze all other application managed by Xpra. > > Even new application started from the tray menu are not responding to > > the mouse. > Unless the server eventually disconnects, this sounds like a modal > window or a grab. > > > if i ssh to the workstation and kill the application everything is start > > working again. I think this particular problem is coming from a bad > > behavior of Rabbitvcs.. I am not a expert in this kind of software but > > the fact that all the windows are frozen seems wrong. > Could be, it could also be a bug in xpra. > > Not sure how to tackle this situation. Not related to java application > > and i have seen this behavior with other perl/tk application, > > dophin file manager etc. > I've just tried it in an Ubuntu 19.10 virtual machine and I couldn't see > any obvious problems. > Do I have to use a specific dialog or action to trigger this bug? > > Assuming that this isn't a modal window problem, would you mind creating > a ticket for this problem? > > Thanks, > Antoine > > > > Carl. > > > > > > > > > > On Fri, Apr 17, 2020 at 2:55 AM Antoine Martin via shifter-users > > > > wrote: > > > > On 17/04/2020 00:35, carl langlois via shifter-users wrote: > > > Hi, > > > > > > Using 3.0.9 on both server(Ubuntu 19.10) and client(windows 10) > > the mouse > > > does not seem to click at the good location. Like if all focus is > > wrong > > > between the windows and the mouse. > > Is this with all applications or just a specific one? For all its > > windows, or maybe just popup menus? > > Does it not work with a plain xterm, or maybe your window is > maximized? > > If the offset large or small? Are you using Wayland client-side > perhaps? > > > > There are a number of tickets related to pointer offsets, but they > > generally involve Java or proprietary applications. Here's a ticket > > which may be related: > > http://xpra.org/trac/ticket/2600 > > If this is the right place, please add your information there. > > > > Cheers, > > Antoine > > > > > > > > > > Thanks and Regards > > > > > > Carl > > > _______________________________________________ > > > shifter-users mailing list > > > shifter-users at lists.devloop.org.uk > > > > > https://lists.devloop.org.uk/mailman/listinfo/shifter-users > > > > > > > _______________________________________________ > > shifter-users mailing list > > shifter-users at lists.devloop.org.uk > > > > https://lists.devloop.org.uk/mailman/listinfo/shifter-users > > > > From crl.langlois at gmail.com Fri Apr 17 17:55:02 2020 From: crl.langlois at gmail.com (carl langlois) Date: Fri, 17 Apr 2020 12:55:02 -0400 Subject: [winswitch] Focus issue In-Reply-To: <1ced558c-b659-2dce-93ad-70fcac650e8c@nagafix.co.uk> References: <120866b1-bf6a-55b0-6381-41c704ecddb3@nagafix.co.uk> <7373749a-1d4f-4624-a73b-d6a780d213b6@nagafix.co.uk> <1ced558c-b659-2dce-93ad-70fcac650e8c@nagafix.co.uk> Message-ID: Thanks for the info i will keep it disable from now on. Yes the two popup windows from rabbitvcs are freezing with full desktop. Thanks again for the quick response. Carl On Fri, Apr 17, 2020 at 12:13 PM Antoine Martin wrote: > On 17/04/2020 23:02, carl langlois wrote: > > HI, > > > > You are right this seem to be a modal issue. Reproducing the problem > > with modal disable fix the freezing issue. > Great. > > FYI: starting with version 4.0, modal-windows will be disabled by default. > > > To trigger this i need to go in the check for modification > > dialog->remote changes->(you need to have some changes) if you double > > click on one of the changes it will popup two windows. On my side the > > two popup are freezing(Rabbitvcs issue ..) > Does it also freeze without xpra? > > > but with modal disable i can > > xkill the rabbitvcs windows and resume normal operation. > > > > This morning i had another phenomena like this one with other perl/tk > > applications but i was able to recover because not all windows was > > affected.. if it happen again i will try disabling the modal windows. > > > > Great hints!! > > Since it is a modal issue to you need me to open a ticket? > No, it doesn't look like one is needed. (or maybe for the rabbitcvs > freeze?) > > Cheers, > Antoine > > > Carl > > > > > > > > On Fri, Apr 17, 2020 at 11:16 AM Antoine Martin > > wrote: > > > > On 17/04/2020 19:26, carl langlois wrote: > > > Hi, > > > When it happen all the windows seems to be affected. It seem that > all > > > the. It look to me that some windows(application) grab the focus > > and all > > > the other application does not received even. > > Please see --modal-windows=no > > A modal window will stay on top of all other windows unless you turn > off > > this feature. > > > I can currently easily reproduce the situation where a windows > > freeze (i > > > suspect a application bugs) . It seem that in this case the > > application > > > enter a weird state and freeze all other application managed by > Xpra. > > > Even new application started from the tray menu are not responding > to > > > the mouse. > > Unless the server eventually disconnects, this sounds like a modal > > window or a grab. > > > > > if i ssh to the workstation and kill the application everything is > > start > > > working again. I think this particular problem is coming from a bad > > > behavior of Rabbitvcs.. I am not a expert in this kind of software > but > > > the fact that all the windows are frozen seems wrong. > > Could be, it could also be a bug in xpra. > > > Not sure how to tackle this situation. Not related to java > application > > > and i have seen this behavior with other perl/tk application, > > > dophin file manager etc. > > I've just tried it in an Ubuntu 19.10 virtual machine and I couldn't > see > > any obvious problems. > > Do I have to use a specific dialog or action to trigger this bug? > > > > Assuming that this isn't a modal window problem, would you mind > creating > > a ticket for this problem? > > > > Thanks, > > Antoine > > > > > > > Carl. > > > > > > > > > > > > > > > On Fri, Apr 17, 2020 at 2:55 AM Antoine Martin via shifter-users > > > > > > > > >> wrote: > > > > > > On 17/04/2020 00:35, carl langlois via shifter-users wrote: > > > > Hi, > > > > > > > > Using 3.0.9 on both server(Ubuntu 19.10) and client(windows > 10) > > > the mouse > > > > does not seem to click at the good location. Like if all > > focus is > > > wrong > > > > between the windows and the mouse. > > > Is this with all applications or just a specific one? For all > its > > > windows, or maybe just popup menus? > > > Does it not work with a plain xterm, or maybe your window is > > maximized? > > > If the offset large or small? Are you using Wayland > > client-side perhaps? > > > > > > There are a number of tickets related to pointer offsets, but > they > > > generally involve Java or proprietary applications. Here's a > > ticket > > > which may be related: > > > http://xpra.org/trac/ticket/2600 > > > If this is the right place, please add your information there. > > > > > > Cheers, > > > Antoine > > > > > > > > > > > > > > Thanks and Regards > > > > > > > > Carl > > > > _______________________________________________ > > > > shifter-users mailing list > > > > shifter-users at lists.devloop.org.uk > > > > > > > > > > > https://lists.devloop.org.uk/mailman/listinfo/shifter-users > > > > > > > > > > _______________________________________________ > > > shifter-users mailing list > > > shifter-users at lists.devloop.org.uk > > > > > > > > > > https://lists.devloop.org.uk/mailman/listinfo/shifter-users > > > > > > > From crl.langlois at gmail.com Wed Apr 22 15:16:50 2020 From: crl.langlois at gmail.com (carl langlois) Date: Wed, 22 Apr 2020 10:16:50 -0400 Subject: [winswitch] Mouse button behavior. Message-ID: Hi, Many time the mouse button behave incorrectly. When this happen the left mouse button seem to do a right click instead of left click. In this situation clicking on the right button will put the situation back to normal. Any hints on what to look for or option to try. to debug this behavior. Thanks and regards. Carl From antoine at nagafix.co.uk Wed Apr 22 15:31:58 2020 From: antoine at nagafix.co.uk (Antoine Martin) Date: Wed, 22 Apr 2020 21:31:58 +0700 Subject: [winswitch] Mouse button behavior. In-Reply-To: References: Message-ID: > Hi, > > Many time the mouse button behave incorrectly. When this happen the left > mouse button seem to do a right click instead of left click. In this > situation clicking on the right button will put the situation back to > normal. Do you have specific steps to reproduce this problem? Is it a regression? Is this using MS Windows as client perhaps? > Any hints on what to look for or option to try. to debug this behavior. This sort of thing is quite hard to debug, you may want to run the server with `xpra start -d mouse` then you can look for button events in the log as `xtest_fake_button`. Button 1 is the left, 3 is right. (True for pressed, False for unpressed) If that doesn't clear things up, please create a ticket with reproducible steps. I should be able to fix that. My guess is that we don't get the button release event for some reason (maybe sent to another window when we don't have focus). Cheers, Antoine > > Thanks and regards. > Carl From crl.langlois at gmail.com Wed Apr 22 18:02:21 2020 From: crl.langlois at gmail.com (carl langlois) Date: Wed, 22 Apr 2020 13:02:21 -0400 Subject: [winswitch] Mouse button behavior. In-Reply-To: References: Message-ID: No i currently do not have specific step to reproduce since it is happening randomly. I will try to see if i can have reproducible step. Yes it is windows client. We have different connection speed and lot of variation in latency. It seem to happen more frequently on bad connection. Your guess seems to be a good one.. I will run the server with the -d mouse and see what happens. Thanks and regards Carl On Wed, Apr 22, 2020 at 10:32 AM Antoine Martin via shifter-users < shifter-users at lists.devloop.org.uk> wrote: > > Hi, > > > > Many time the mouse button behave incorrectly. When this happen the left > > mouse button seem to do a right click instead of left click. In this > > situation clicking on the right button will put the situation back to > > normal. > Do you have specific steps to reproduce this problem? > Is it a regression? > Is this using MS Windows as client perhaps? > > > Any hints on what to look for or option to try. to debug this behavior. > This sort of thing is quite hard to debug, you may want to run the > server with `xpra start -d mouse` then you can look for button events in > the log as `xtest_fake_button`. > Button 1 is the left, 3 is right. (True for pressed, False for unpressed) > > If that doesn't clear things up, please create a ticket with > reproducible steps. I should be able to fix that. > My guess is that we don't get the button release event for some reason > (maybe sent to another window when we don't have focus). > > Cheers, > Antoine > > > > > Thanks and regards. > > Carl > _______________________________________________ > shifter-users mailing list > shifter-users at lists.devloop.org.uk > https://lists.devloop.org.uk/mailman/listinfo/shifter-users > From dragilequarentie at gmail.com Sat Apr 25 22:59:29 2020 From: dragilequarentie at gmail.com (Dr Agile) Date: Sun, 26 Apr 2020 00:59:29 +0300 Subject: [winswitch] nat2nat end-to-end solution using Xpra ? Message-ID: I want to connect to a computer behind NAT. Is there an end-to-end guide / solution using Xpra? Possibly using a self-hosted broker, on my own server. I'm new here, and can't find my way through all the possibilities. From antoine at nagafix.co.uk Sun Apr 26 16:39:31 2020 From: antoine at nagafix.co.uk (Antoine Martin) Date: Sun, 26 Apr 2020 22:39:31 +0700 Subject: [winswitch] nat2nat end-to-end solution using Xpra ? In-Reply-To: References: Message-ID: <35472720-dc6e-c899-6091-3e71e15c4ba8@nagafix.co.uk> On 26/04/2020 04:59, Dr Agile via shifter-users wrote: > I want to connect to a computer behind NAT. > > Is there an end-to-end guide / solution using Xpra? > > Possibly using a self-hosted broker, on my own server. > I'm new here, and can't find my way through all the possibilities. That's probably because there are many ways of doing that. Until xpra supports NAT traversal: https://xpra.org/trac/ticket/2722 most will revolve around using SSH, VPN or netcat to bounce the traffic off your server. Simple example (untested) using TCP mode tunneled via SSH: * start xpra on your "computer behind NAT", ie: xpra start --start=xterm --bind-tcp=0.0.0.0:10000 * create a back tunnel from that computer to your "server" ssh -R 20000:localhost:10000 theserver (keeping this ssh session alive is left as an exercise) * create a tunnel from your client PC to your server: ssh -L 30000:localhost:20000 theserver * then you can connect over those tunnels with: xpra attach tcp://localhost:30000/ I've used different port numbers to try to make it clearer what connects where - though that's not actually required, you could just use the same port number everywhere. Another option would be to forward your computer's SSH server port to your server. Then you could connect "directly" to your computer using an ssh proxy command, or using xpra's "2hop ssh" mode which is available from the connect dialog. Cheers, Antoine